Transcriptomic Studies of the Effect of nod Gene-Inducing Molecules in Rhizobia: Different Weapons, One Purpose.

Simultaneous quantification of transcripts of the whole bacterial genome allows the analysis of the global transcriptional response under changing conditions. RNA-seq and microarrays are the most used techniques to measure these transcriptomic changes, and both complement each other in transcriptome profiling. In this review, we exhaustively compiled the symbiosis-related transcriptomic reports (microarrays and RNA sequencing) carried out hitherto in rhizobia. This review is specially focused on transcriptomic changes that takes place when five rhizobial species, Bradyrhizobium japonicum (=diazoefficiens) USDA 110, Rhizobium leguminosarum biovar viciae 3841, Rhizobium tropici CIAT 899, Sinorhizobium (=Ensifer) meliloti 1021 and S. fredii HH103, recognize inducing flavonoids, plant-exuded phenolic compounds that activate the biosynthesis and export of Nod factors (NF) in all analysed rhizobia. Interestingly, our global transcriptomic comparison also indicates that each rhizobial species possesses its own arsenal of molecular weapons accompanying the set of NF in order to establish a successful interaction with host legumes.

1. Introduction

Quantification of the global expression level of a bacterial genome in response to specific cues allows the analysis of the transcriptional responses to changing environmental conditions. Several transcriptome profiling methods have been developed in the last years. DNA microarrays and RNA sequencing (RNA-seq) are the two most used genome-wide gene expression quantification techniques [1,2,3,4].

Microarrays are based on the hybridization of oligonucleotide DNA sequences representing the entire set of genes of an organism arranged in a grid pattern with DNA molecules representing the transcriptome in a specific condition [5,6,7]. In contrast, in the case of the RNA-seq methodology, the whole complementary DNA (cDNA) is directly and massively sequenced. By mapping the resulting short-sequence reads onto the reference genome, both the relative and absolute expression levels of each gene can be quantified [3,8]. RNA-seq technology does not require species- or transcript-specific probes and, in comparison with microarrays, offers increased specificity and sensitivity. Interestingly, alternative transcripts, small untranslated RNA and transcription start sites can also be identified by using this technique [9,10]. However, the efficiency of RNA-seq is marred with the problem of an overwhelming amount of ribosomal RNA (rRNA) in the data (although rRNA can be almost totally removed during the library preparation), and with variations of reads density along the length of the transcript (normalization of reads density variation along the length of the transcript solves this problem) [11]. For these reasons, the application of multiple transcriptome profiling methods improves the comprehension of the global gene expression picture rather than basing only on one method.

Rhizobia have two distinct life styles, either as free-living soil bacteria or as nitrogen-fixing endosymbionts, called bacteroids, within root nodules of legume host plants. Differentiation from free-living bacteria to bacteroids requires important transcriptomic changes in rhizobia, since bacteria adapt their metabolism to nitrogen fixation and to living inside nodule plant cells. In the case of nodules formed by IRLC-legumes (Inverted Repeat-Lacking Clade), this differentiation process is more drastic and leads to the formation of endoreduplicated bacterial cells that have lost their capacity to divide [12,13]. These bacteroids reduce atmospheric nitrogen to ammonia, which is assimilated by the host plant in exchange of a carbon source and an appropriate environment that promotes the growth of undifferentiated rhizobia [14,15,16]. The oxygen limitation appears to be the key factor driving expression of nitrogen-fixation genes inside root nodules and the nif and fix cluster of genes are strongly upregulated under microaerobic conditions [17,18,19,20]. However, before this differentiation process occurs, other molecular signals play crucial roles in the phenotypic changes required for the transition between free-living rhizobia and bacteroids.

The initial steps of the symbiotic interaction between legumes and nitrogen-fixing rhizobia involve reciprocal communication by means of chemical signals produced by both members of the symbiosis. This specific interaction starts when the plant releases to the rhizosphere flavonoids that are recognized by a Lysine transcriptional Regulator (LysR)-type in the bacterium, the NodD protein. NodD induces the expression of the nodulation (nod, nol, noe) genes by binding to specific sequences, the nod boxes (NB), located upstream of these genes [21]. These nodulation genes code for enzymes implied in the biosynthesis and export of lipochitooligosaccharides, acylated chitin oligomeric backbones with various functional group substitutions at the terminal or non-terminal residues, also known as Nod factors (NF). These signal molecules are recognized by a specific class of receptor kinases, called Lysin Motif Receptor-Like Kinase (LysM-RLK), which harbour a lysine motif in their extracellular domains. Two LysM receptor kinases (Nod-Factor Receptors, NFR1 and NFR5) are present in roots of different legumes such as the model legumes Lotus japonicus and Medicago truncatula [22]. Nod factors induce nodule primordia in plant roots and play an essential role in the infection process [23,24,25]. Interestingly, many other symbiotic-related traits are regulated in the first instance via flavonoids and the NodD protein in some rhizobia, such as production of surface polysaccharides, synthesis of phytohormone, motility, quorum sensing, and activation of the type 3 secretion system (T3SS) [26,27,28,29,30,31,32]. However, these symbiotic-related features are not present in all rhizobia, since each microsymbiont has developed its own molecular arsenal to reach the same purpose: the success in the infection of host legumes.

In this review, we extensively searched and analysed the main symbiotic-related transcriptomic studies carried out hitherto in rhizobia by using microarray and/or RNA-seq techniques. Our study is specifically focused on those reports in which it was analysed the global gene expression response of several rhizobial strains to the presence of the key initial symbiotic signals, the inducing flavonoids. The goal of this review is to describe the arsenal of molecular weapons used by each rhizobial strain in the initial steps of its symbiotic interactions with host legumes.

2. Rhizobial Transcriptomic Studies Related with Symbiosis

As previously discussed, the rhizobium-legume interaction is characterized by a complex exchange of signal molecules from both partners that culminates in the formation of nodules, in which biological nitrogen fixation takes place. Two of the most important and studied conditions are the presence of flavonoids (which induce nod genes expression) and a microaerobic environment (which triggers the expression of the nif/fix genes). In fact, the transcriptomic reports published during the last 15 years can be grouped depending on the specific stages of the symbiotic process analysed: bacteroids vs. free-living bacteria, microaerobiosis vs. aerobiosis, and rhizosphere or presence of root exudates or flavonoids vs. non-induced cultures (Figure 1, Table 1).

Figure 1

Transcriptomic conditions analysed in this review. Reports compiled in this manuscript were performed in different stages/conditions of the symbiotic process (flavonoids, root secretions, rhizosphere, microaerobiosis and bacteroids).

Table 1

Summary of the main results obtained in rhizobial transcriptomic studies related with the symbiotic process.

Comparison	Bacteria	Method	Strains	Specific Conditions	Results Related with the Symbiotic Process	References
Free-living cells vs. Bacteroids	Bradyrhizobium japonicum USDA 110	Microarray (AffymetrixGeneChip)	Wild-type and RNA polymerase σ54 factor (rpoN1-2) double mutant	-Free-living cultures in either aerobic or microaerobic conditions (mid-exponential phase).-Bacteroids in soybean (10, 13, 21 and 31 days post inoculation, dpi).	-Distinction among genes expressed in early and late bacteroids.-1/3 of the genes induced in bacteroids (21 dpi) are also upregulated in microaerobiosis, including fix and nif genes. The other induced genes are regulated by factors other than O2 limitation.-Determination of genes transcribed in bacteroids in a σ54 (rpoN1)-dependent manner (several fix, nif and hup genes included).	[19]
	B. japonicum USDA 110	Microarray(Operon Biotechnologies)	Wild-type	-Minimal and rich media (mid-exponential phase).-Osmotic stress conditions (mid-exponential phase).-Bacteroids in soybean (28 dpi).	-15% of the genes of the genome were differentially expressed in bacteroids: N2 fixation (nif and fix) and H2 uptake (hup) genes up-regulated. Nodulation (nod) genes also up-regulated. T3SS (hrc) genes are repressed.	[33]
	Rhizobium leguminosarum biovar viciae 3841	Microarray(Operon Biotechnologies)	Wild-type	-Several carbon substrates (glucose, pyruvate, succinate, inositol, acetate, and acetoacetate).-Bacteroids in vetch (28 dpi).-Bacteroids in pea (7, 15, 21 and 28 dpi).	-386 genes were differentially expressed in at least one stage of bacteroid development.-The decarboxylating arm of the tricarboxylic acid cycle and the aminobutyrate metabolism were highly induced in early nodule bacteroids (7 dpi).-Changes in the expression of regulators, exported and cell surface molecules, multidrug exporters, and heat and cold shock proteins in bacteroids of early pea nodules (7 dpi). -The fix genes were induced early but continued to increase in mature bacteroids (15 and 21 dpi), while nif genes were induced strongly in older bacteroids of pea nodules (28 dpi).	[34]
	Mesorhizobium huakuii 7653R	RNA-SeqMicroarray(Roche NimbleGen)	Wild-type	-Bacteroids in Astragalus sinicus (32 dpi).	-Mostly, bacteroid up-regulated genes are located on plasmid while downregulated genes are chromosomal.-Free-living cells have a primary role in maintaining basal metabolism, whereas bacteroids in N2 fixation (nif, fix and suf clusters up-regulated).-Up-regulation of rpoN (σ54) and rpoH (RNA polymerase σ32 factor) indicates an essential role of both transcriptional factors in the activation of genes implied in the N2 fixation.	[35]
	Mesorhizobium loti MAFF303099	Microarray(Made by Authors)	Wild-type	-Free-living cultures in either aerobic or microaerobic conditions (mid-exponential phase).-Bacteroids in Lotus japonicus (42 dpi).-C-starving conditions.	-Clusters of genes within the symbiotic islands are collectively up-regulated in bacteroids (including nif, fix, fdx, and rpoN genes) through σ54 (rpoN).-Most of the upregulated genes in bacteroids are induced in a NifA-independent manner.-Transcription outside the symbiotic island is strongly repressed in bacteroids (genes implied in cell division, cell wall and flagella synthesis).	[20]
	Sinorhizobium meliloti 1021	Macroarrays(Made by authors)	Wild-type and bacA mutant	-Free-living cultures (minimal and rich media) with and without luteolin.-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids in young (8 dpi) nodules and in mature nitrogen-fixing nodules (18 dpi) in Medicago truncatula. -Bacteroids in Medicago sativa mature nodules (18 dpi).	-15 out of 214 tested genes were induced in young and mature nodules of the two strains (specific of the infection stage), including genes that code for hemolysin calcium-binding proteins, adenylate cyclases and type-IV secretion system proteins.-24 out of 214 tested genes were up-regulated in mature nodules of both plants (specific of symbiosis), including nif, fix and cya genes.-18 out of 214 tested genes were down-regulated in mature nodules of both plants (specific of symbiosis), including genes implied in cell division (ftsZ) and chaperonins.	[36]
	S. meliloti 1021	Microarray and Macroarray(Made by authors)	Wild-type	-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids from M. sativa nodules (18 to 22 dpi).	-982 genes differentially expressed in bacteroids.-Genes downregulated were implied in cell division (ftsZ), DNA, RNA and protein metabolisms (dna and rpo cluster genes), motility and chemotaxis (che genes), phosphorus uptake and utilization (pho genes), nitrogen assimilation (gln genes) and glucolysis and aerobic respiration (nuoACEFHIKLMN operon).-Genes upregulated included mainly those related with symbiosis (several nod and noe genes), nitrogen fixation (nif and fix operons) and transport of peptides and aminoacids.	[18]
	S. meliloti 1021	Microarray(Made by authors)	Wild-type and bacA mutant	-Free-living cultures in early and late exponential and stationary phase.-Bacteroids from M. sativa nodules at 5, 8, 14–18 dpi in the wild-type.-Bacteroids from M. sativa nodules at 11 dpi in the mutant.	-36 genes were specifically induced in early stages of the symbiosis (nod genes and genes implied in the synthesis of the cytochrome c).-Bacterial transcriptomic profile changes during nodule development. -The nif and fix cluster genes were up-regulated even in young nodules.	[37]
	S. meliloti 1021	Microarray (AffymetrixGeneChip)	Wild-type, triple nodD1 nodD2 nodD3 mutant, rpoN and fixJ mutants, and triple nodD1 nodD2 nodD3 mutant overexpressing either nodD1 or nodD3	-Minimal and rich media (mid-exponential phase).-Late exponential cultures diluted to 0.15–0.2 OD600 and induced for 4 h with luteolin.-Bacteroids from M. truncatula nodules at 33–35 dpi.	-RpoN controlled genes that are not differentially expressed in free-living bacteria (fdx, nif and fix operons).-Study of gene expression simultaneously in both symbiotic partners indicate that in total more than 5000 genes are differentially expressed in both organisms.-Most of the plant genes upregulated in wild-type nodules were also induced in Fix− nodules (fixJ mutant), indicating that the nitrogen fixing status does not significantly affect plant transcriptomic changes.	[38]
	S. meliloti 1021	Microarray (Made by authors)	Wild-type and and fixJ, nifA, fixK, and nifH mutants	-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids from M. truncatula nodules at 14 dpi.	-122 genes were activated by FixJ via NifA and FixK (97% located in the symbiotic plasmid), including fix and nif genes. -FixJ controls the majority of genes expressed in mature bacteroids.-NifA activated genes implied in nitrogen fixation and exopolysaccharides (EPS) production and FixK up-regulated genes involved in respiration, arginine metabolism, denitrification and stress responses.	[39]
	S. meliloti 2011	RNA-seq(Illumina Hiseq 2000)	Wild-type and rpoE2 mutant	-Free-living cultures in mid-exponential or early stationary phase.-Bacteroids from M. truncatula nodules at 10 dpi.	-Authors designed EuGene-P, a tool that enables the automated prediction of coding sequences, untranslated regions, transcription start sites and non-coding RNA genes.-Prediction of 6308 coding sequences, 1876 non-coding RNAs and 4077 transcription start sites using cDNA of all conditions.	[10]
	S. meliloti 2011	RNA-seq(Illumina Hiseq 2000)	Wild-type	-Bacteroids from M. truncatula nodules at 10 and 15 dpi.	-Coupling bacterial and plant gene transcriptome determination to laser dissection, authors determined expression changes at the tissue level on indeterminate nodules.	[40]
	Sinorhizobium fredii NGR234	RNA-seq(Illumina Hiseq 2000)	Wild-type	-Bacteroids from Vigna unguiculata nodules (21 dpi nodules) -Bacteroids from Leucaena leucocephala (35 dpi nodules).	-3143 genes differentially expressed in bacteroids of V. unguiculata and 2780 in bacteroids of L. leucocephala.-Upregulated genes in bacteroids from both hosts were implied in the synthesis of ATP-binding cassette (ABC) transporters, type 3 secretion system (T3SS), nitrogen metabolism, nitrogen fixation (nif and fix operons), fatty acid metabolism, benzoate degradation and exopolysaccharide biosynthesis (exo cluster genes).-Downregulated genes in bacteroids of both plants were involved in synthesis of DNA, RNA and protein metabolisms and flagellar assembly.	[41]
Aerobiosis (Free-living cells)vs.Microaerobiosis	B. japonicum USDA 110	Microarray (AffymetrixGeneChip)	Wild-type and rpoN1-2 double mutant	-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids in soybean (10, 13, 21 and 31 dpi).	-Microaerobiosis triggered upregulation of symbiotic relevant genes, including fix and nif clusters, mostly in a σ54-dependent manner.-1/3 of the genes induced in bacteroids (21 dpi) are also upregulated in microaerobic conditions.	[19]
	M. loti MAFF303099	Microarray(Made by authors)	Wild-type	-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids in L. japonicus MG20 (42 dpi).-C-starved cells.	-The genome region containing the fixNOPQ genes (outside the symbiosis island) and the fix and nif regions (in the symbiotic island) were upregulated as under both microaerobic and symbiotic conditions.	[20]
	S. meliloti 021	Macroarrays(Made by authors)	Wild-type and bacA mutant	-Free-living cultures (minimal and rich media) with and without luteolin.-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids in young (8 dpi) nodules and in mature nitrogen-fixing nodules (18 dpi) in M. truncatula J6. -Bacteroids in M. sativa mature nodules (18 dpi).	-8 out of 214 tested genes were found to be induced under microoxic and bacteroids conditions, including fix and nif genes.	[36]
	S. meliloti 1021	Microarray (Made by authors)	Wild-type and and fixJ, nifA, fixK, and nifH mutants	-Free-living cultures in either aerobic or microaerobic conditions (mid-exponential phase).-Bacteroids from M. truncatula nodules at 14 dpi.	-FixJ controlled 74% of the genes induced in microaerobiosis and the majority of genes expressed in mature bacteroids (including fix and nif genes).	[39]
	S. meliloti 1021	Microarray and Macroarray(Made by authors)	Wild-type	-Free-living cultures in either aerobic or microaerobic conditions.-Bacteroids from M. sativa nodules (18 to 22 dpi).	-377 genes regulated by oxygen concentration (266 induced and 111 repressed in microaerobic conditions).-31 genes were induced both under microoxic and bacteroids conditions, including genes implied in N2-fixation (nif and fix operons), proline metabolism and denitrification.	[18]
Free-living cells vs.Rhizosphere	R. leguminosarum biovar viciae 3841	Microarray(Operon Biotechnologies)	Wild-type and mutants in many different genes	-Free-living cells vs. rhizosphere (pea, alfalfa or sugar beet) attached bacteria (7 dpi).-Free-living cultures with and without root exudates (pea).-Free-living cultures with and without the flavonoid hesperetin.	-A common core of 106 genes were rhizosphere-induced genes (70 genes encode for proteins with unknown functions).-The increase of gene expression of the glyoxylate cycle only occurred in the pea rhizosphere.-Many genes on pRL8 (plasmid 8 of R. leguminosarum) were specifically up-regulated in the pea rhizosphere.	[42]
Non-induced (Free-living cells)vs.root secretions	R. leguminosarum biovar viciae 3841	Microarray(Operon Biotechnologies)	Wild-type and mutants in many different genes	-Free living cells vs. rhizosphere (pea, alfalfa or sugar beet) attached bacteria (7 dpi).-Free-living cultures with and without root exudates (pea).-Free-living cultureswith and without the flavonoid hesperetin.	-21 genes up-regulated including the nodulation (nod) and rhi gene clusters on pRL10 (the symbiotic plasmid).	[42]
Non-induced (Free-living cells)vs.flavonoids	B. japonicum USDA 110	Microarray (AffymetrixGeneChip)	Wild-type,nodW mutant and nodW mutant overexpressing nwsB	-Free-living cultures with and without genistein (8 hpi).	-101 genes up-regulated in the presence of genistein, including nod genes, the flagellar cluster and transport genes.-NodW was essencial for induction of most of these genes.-The phenotype and the gene expression levels in the nodW mutant were partially restored by overexpression of nwsB gene.	[43]
	R. leguminosarum biovar viciae 3841	Microarray(Operon Biotechnologies)	Wild-type and mutants in many different genes	-Free-living cells vs. rhizosphere (pea, alfalfa or sugar beet) attached bacteria (7 dpi).-Free-living cultures with and without root exudates (pea).-Free-living cultures with and without the flavonoid hesperetin.	-27 genes up-regulated in the presence of flavonoids, including the nod and rhi gene clusters on pRL10 (the symbiotic plasmid).-6 genes down-regulated, including flaD.	[42]
	Rhizobium tropici CIAT 899	RNA-seq(Illumina Hiseq 2000)	Wild-type,nodD1 mutant andnodD2 mutant	-Free-living cultures with and without the flavonoid apigenin.-Free-living cultures with and without salt.	-13 symbiotic-related genes up-regulated in the presence of apigenin, including nod genes and the IAA synthesis genes.-2 hypothetical-protein genes putatively related with symbiosis were up-regulated via NodD1 and apigenin.-NodD1 activated the expression of the 13 symbiotic-related found genes.	[44,45]
	S. meliloti 1021	Macroarrays(Made by authors)	Wild-type and bacA mutant	-Free-living cultures (minimal and rich media) with and without the flavonoid luteolin.-Aerobic and microaerobic conditions.-Bacteroids in young (8 dpi) nodules and in mature nitrogen-fixing nodules (18 dpi) in M. truncatula J6. -Bacteroids in M. sativa mature nodules (18 dpi).	-7 out of the 214 tested genes were induced with luteolin, including the nod genes, traA and three genes related to iron metabolism (also in nodules).	[36]
	S. meliloti 1021	Microarray (Made by authors)	Wild-type and wild-type overexpressing nodD1	-Late exponential cultures diluted to 0.15–0.2 OD600 and induced for 4 and 24 h with luteolin.	-26 and 5 genes after 4 and 24 h, respectively, were affected in the presence of genistein, including those belonging to nod/nol/noe operons. -Other genes encoding for hyphotetical proteins were also identified (varying among replicates).	[46]
	S. meliloti 1021	Microarray (AffymetrixGeneChip)	Wild-type, triple nodD1 nodD2 nodD3 mutant, rpoN and fixJ mutants, and triple nodD1 nodD2 nodD3 mutant overexpressing either nodD1 or nodD3	-Minimal and rich media (mid-exponential phase).-Late exponential cultures diluted to 0.15–0.2 OD600 and induced for 4 h with luteolin.-Bacteroids from M. truncatula nodules at 33–35 dpi.	-Luteolin induced significant expression changes in nod/nol/noe operons (12 genes).-Other 12 genes did not show true flavonoid induction but instead vary in expression depending on replicate.	[38]
	Sinorhizobium fredii HH103	RNA-seq(Illumina Hiseq 2000)	Wild-type, nodD1 and ttsI mutants	-Free-living cultures (early stationary phase) with and without the flavonoid genistein.	-100 genes were affected in the presence of genistein: 70 genes induced through nod boxes (nod, nol and noe genes) and tts boxes (T3SS genes) were upregulated. 30 genes not controlled by NB or TB were differentially expressed in the presence of genistein.	[47]
Non-induced (Free-living cells)vs.saline stress (nod gene inducing conditions)	R. tropici CIAT 899	RNA-seq(Illumina Hiseq 2000)	Wild-type, nodD1 mutant andnodD2 mutant	-Free-living cultures with and without the flavonoid apigenin.-Free-living cultures with and without salt.	-17 symbiotic-related genes up-regulated in the presence of salt, including nod genes and the IAA synthesis genes.-All the symbiotic-related genes upregulated in presence of apigenin are upregulated in salt conditions as well.-2 hypothetical-protein genes putatively related with symbiosis were upregulated via NodD2 and salt.-In general, higher nod gene expression was detected with salt than with apigenin.-NodD2 activated the expression of the 17 symbiotic-related found genes.-NodD1 enhanced the expression of nodD2 under salt conditions.	[44,45]

RNA-seq: RNA sequencing; OD600: optical density at 600 nm, NB: nod boxes; TB: tts boxes.

2.1. Bacteroid and Microaerobiosis Conditions

Microarray and/or RNA-seq studies performed in bacteroids or in microaerobiosis conditions has allowed the analysis of the global transcriptomic changes necessary for the reduction of atmospheric nitrogen to ammonia in the symbiosomes of nodules (Table 1). It has been previously described that two Nif and one Fix proteins are the components of the symbiotic nitrogen fixation machinery. The nif genes code for the two proteins that constitute the nitrogenase (dinitrogenase and dinitrogenase reductase) as well as for the iron-molybdenum cofactor synthesis proteins, whereas ferredoxins or flavoproteins, the electron supply proteins, are encoded by the fix genes [48]. These two main clusters of genes are directly responsible for N2 fixation and their expression is controlled primarily by the oxygen tension in rhizobia, although nitrogen availability also affects their expression levels [49,50]. Thus, under low O2 tension, the oxygen sensor FixL autophosphorylates, and transfers phosphate to its cognate response regulator FixJ, which in turn up-regulates the transcription of the nifA and fixK genes. NifA, in conjunction with RpoN (σ54 transcriptional factor) activates the transcription of nitrogen fixation genes, such as nifHDKE and fixABCX operons. Moreover, FixK also induces the transcription of other nitrogen fixation genes, such as the fixNOQP and fixGHIS operons.

Interestingly, this hierarchical fixJ-nifA-fixK regulatory cascade was fully assessed by performing transcriptomic studies in Sinorhozobium meliloti 1021 [39]. The rest of transcriptomic studies confirmed that both fix and nif clusters are strongly up-regulated in symbiosomes in all analysed rhizobia (Table 1). Interestingly, in Bradyrhizobium japonicum and Mesorhizobium spp., other groups of genes related with the correct functioning of the nitrogenase enzyme, such as the hup (H2 uptake), fdx (ferredoxin) and suf (Fe-S cluster) operons, are commonly activated in symbiosomes [20,33,35]. The transcriptomic studies performed with rpoN mutants under low O2 concentration conditions indicate that the upregulation of N2 fixation-related clusters is mediated by oxygen limitation and the σ54 transcriptional factor (RpoN) in Bradyrhizobium japonicum USDA 110, Mesorhizobium huakuii 7653R and Sinorhizobium meliloti 1021 [19,33,35,38]. However, in the case of M. loti MAFF303099, the transcriptomic studies performed in parallel under microaerobiosis conditions indicated that most of the up-regulated genes in bacteroids are induced in a NifA-RpoN-independent manner [20]. On the other hand, transcriptomic studies carried out in symbiosomes of Sinorhizobium spp. indicate that several processes such as cell division (ftsZ gene), DNA, RNA and protein metabolisms are inhibited, which might be correlated with the absence of cell division in fully developed bacteroids [18,37,41].

Lastly, the paper published by Roux et al. [40] entails a milestone in the field of transcriptomic research in symbiosis since authors studied for the first time how bacterial and plant gene expression levels vary in different tissues of the indeterminate nodules formed in the symbiosis between S. meliloti and M. truncatula.

2.2. Rhizosphere and Exudates Conditions

Flavonoids are phenylpropanoid metabolites exuded by plant roots, often in response to elicitors produced by microorganism present in the rhizosphere [51]. These molecules are antimicrobial compounds that inhibit growth of many root pathogens, ranging from bacteria to fungi [52]. However, in rhizobia, flavonoids can act as bacterial growth promoters, chemoattractants, inducers of nodulation genes, determinants of host specificity, and regulators of phytoalexin resistance [53]. Transcriptomic studies performed in rhizosphere conditions or in culture media supplemented with either root exudates or flavonoids have shed light to the repertoire of molecular weapons used by each rhizobial strain in the initial steps of the symbiosis (Table 1). However, only one study has been developed in the case of transcriptomic changes that take place in the rhizosphere and in the presence of root exudates of different host and non-host legumes as well as of non-legume plants. Results obtained by Ramachandran et al. [42] described a common core of rhizosphere-induced genes of Rhizobium leguminosarum biovar viciae 3841. Despite the fact that most of these genes code for proteins with unknown functions (66%), the up-regulation of well-known genes indicates that several physiological processes are enhanced when this rhizobial species colonizes plant roots, such as acid organic metabolism (dctA and pckA genes, implied in C4-dicarboxylate transport and Phosphoenolyruvate (PEP)-carboxykinase production, respectively), molecular transport (rmrA, coding for a multi-drug resistance family efflux pump) [54] and osmotic adaptation (ndvA, whose product is responsible for the export of cyclic β-1-2-glucans to the bacterial periplasm) [55]. The activation of this set of genes appears to be the way R. leguminosarum deals with adverse conditions in the plant rhizosphere.

Interestingly, many Rhizobium leguminosarum bv. viciae (Rlv)3841 genes are induced in the rhizosphere of legume plants, including the nodABCEFIJLMNO operon (located on pRL10, the symbiotic plasmid of Rlv3841), but not in the sugar beet rhizosphere. As expected, this nod operon is also induced in the presence of pea exudates, containing inducing flavonoids for this strain. Surprisingly, many genes of the non-symbiotic plasmid pRL8 are upregulated only in the rhizosphere of pea, which indicates that pRL8 enables adaptation of R. leguminosarum to this specific environment. Among these genes are included those coding for a molybdenum-containing xanthine dehydrogenase-like carbon monoxide dehydrogenase (cox cluster) and proteins involved in molybdopterin biosynthesis (moaA gene) [42]. However, the exact role of these proteins in the adaptation of R. leguminosarum to pea rhizosphere remains to be studied.

3. Gene Expression Changes under nod Gene-Inducing Molecules: The Arsenal of Weapons

The symbiotic relationship between legumes and nitrogen-fixing rhizobia involves reciprocal communication through chemical signals produced by the plant and the bacterium. The key plant molecules that initiate this communication are flavonoids. It is well known that these inducers activate the expression of the nod genes, which code for enzymes involved in the production and export of specific NF that are responsible for root infection and induction of nodule development [53]. However, upregulation of nod genes is only the tip of the iceberg in the transcriptomic changes that take place when rhizobia perceive the appropriate flavonoids. Proteomic studies also support this observation since, besides Nod proteins, many other rhizobial proteins are differentially synthetized in the presence of inducing flavonoids [56,57].

3.1. Bradyrhizobium Japonicum USDA 110: Diversification in the Activation Pathways for NF Production, Assembly of the T3SS and the Flagellum

B. japonicum USDA 110 is an endosymbiont of soybean (Glycine max), mung bean (Vigna radiata), cowpea (Vigna unguiculata) and siratro (Macroptilium atropurpureum), among other legumes [58]. In this bacterium, the regulation of symbiotic genes differs from that of other rhizobial strains. Thus, besides NodD1, the two-component regulator NodW, which is phosphorylated by the NodV sensor protein, promotes the transcription of nod genes [59,60,61,62,63]. Interestingly, nodW seems to play a major role in this strain, since its mutation prevents nodulation in mung bean, cowpea and siratro [59]. In contrast, mutation in nodD1 causes severe effects in the nodulation with mung bean and siratro, but does not completely block the nodulation process [63]. In both mutants, nodulation with soybean is less efficient, leading to a delay in the formation of nodules [59,60,61,62,63]. Surprisingly, the symbiotic defect of the nodW mutant can be suppressed by overexpression of the gene coding for another two-component regulator, nwsB [64].

An analysis of B. japonicum USDA 110 transcriptome changes in response to genistein, an inducing flavonoid for this strain, was performed using the wild type strain, a nodW mutant and this mutant overexpressing nwsB [43]. This work showed that this bacterium responds to genistein by triggering the expression of a battery of 101 genes after 8 h of treatment with this flavonoid (Table S1). Interestingly, three groups of genes differentially expressed were identified: (i) genes controlled by nod boxes (NB); (ii) genes regulated by tts boxes (TB); and (iii) genes not preceded by a NB or a TB (Figure 2).

Figure 2

Transcriptomic response in the presence of genistein in B. japonicum USDA 110. Proteins are illustrated by blue circles. Protein phosphorylation processes are indicated by blue arrows. The DNA to protein translation process is represented by dotted arrows. Activation and repression of genes are indicated by black and red arrows, respectively. Genes regulated by a nod box or a tts box are represented in green and purple, respectively. Genes which are not controlled by a nod box or a tts box are indicated in grey. NB: nod box; TB: tts box; T3SS: type 3 secretion system.

3.1.1. Genes Controlled by NB

Among genes induced with genistein, 34 are located within the symbiotic region, being 20 of them dependent on nod boxes. As expected, this transcriptomic study demonstrated that NodW is the main regulator in the activation of these 20 genes via NB, despite the fact that the interaction of NodW and NodD1 at NB promoters remains to be clarified. In the absence of NodW, the overexpression of NwsB complements the activation pathway via NB [43]. Previous works described that three sets of genes are located downstream of NB, the nodYABCSUIJnolMNnodZ and nolYZ operons [65,66] and the ttsI gene, the latter coding for the transcriptional activator of T3SS genes [67]. As expected, these genes are strongly up-regulated in a NodW-NwsB dependent manner. In addition, by using sequence comparisons and hybridization experiments, another potential NB has been described in the promoter region of the bsr1863 gene, which codes for a hypothetical protein of unknown function [68,69]. The expression of this gene is also controlled by NodW and NwsB (Table S1) (Figure 2).

3.1.2. Genes Controlled by TB

The TtsI protein is a transcriptional regulator that binds to conserved sequences located in the promoter region of the T3SS genes, the tts boxes (TB). In the transcriptomic study performed by Lang et al. [43], only 6 out of the 35 TB previously described [67,68,69,70] are activated eight hours upon induction with genistein. Thus, the expression of the gunA2 (blr1656), blr1676, bsl18089, nopB (blr1812), pgl (blr1993), and blr2140 genes is triggered in the presence of flavonoid in a NodW-dependent manner. Although nopB is the first gene of an T3SS operon, none of the genes located downstream were activated with genistein in this study, which suggests that probably at the time of sample collection (8 h upon addition of genistein) activation of the transcription of the T3SS genes was only starting. In fact, studies performed previously by using fusions to a reporter gene determined that all T3SS genes are induced with genistein [67]. Furthermore, proteomic studies previously performed also support this fact, since in the presence of genistein different T3SS proteins, such as Blr1649, Blr1806 and GunA2 (Blr1656), are secreted to the extracellular medium [70], although only gunA2 was detected as activated in the presence of genistein (Table S1) in this transcriptomic study [43].

3.1.3. Genes Not Preceded by a NB or a TB

Most of these genes belong to the flagellar cluster, which could be related to the competitive capacity of the bacterium in the rhizosphere. All these genes are expressed uniformly and depend on NodW. Interestingly, overexpression of nwsB is not able to restore expression levels in a nodW mutant, which indicates that NwsB complement only genes controlled by NB. Some of the genes that are expressed upon genistein induction code for regulator proteins that could be controlling the expression of the flagellar cluster. Interestingly, Blr6843 and Blr6886 belong to the Multiple antibiotic resistance Regulator (MarR) family of regulators, which are repressor and/or activator proteins that sense environmental changes and may even interact with phenolic compounds [71,72].

On the other hand, the differential expression in the presence of genistein of eight genes located outside of the symbiotic region is not due to NodW [43]. Seven of these genes are related to transport processes (blr2422, blr2423, bll4319, bll4320, bll4321, bll6621 and bll6622 genes): Bll6621 and Bll6622 belong to the major facilitator superfamily, while the remaining proteins belong to the resistance-nodulation-division (RND) family. The presence of Tetracyclin Repressor (TetR)-type repressors close to these genes suggests that their expression could be regulated by these repressors [73]. On the other hand, one of the genes that is not regulated by NodW, blr4684, codes for an esterase with a patatin type domain present in several pathogen effectors but with unknown function in symbiosis. Interestingly, there are homologues to blr4684 in other rhizobial strains, such as Bradyrhizobium sp. BTAil, Sinorhizobium fredii NGR234 and Mesorhizobium loti MAFF303099.

Finally, the nodD1 gene is included in the set of genes not preceded by NB or TB but upregulated by flavonoids and NodW. Interestingly, the expression pattern of nodD1 is restored by NwsB in the absence of NodW. This expression pattern is also observed in other symbiotic genes, such as bll1630 (nolK), bll1631 (noeL) and blr2062 (noeI) genes, involved in 2-O-methyl-fucosylation of NF [61] (Table S1) (Figure 2).

3.2. Rhizobium Leguminosarum Biovar Viciae 3841: NF Synthesis and Adaptation to the Rhizosphere Environment

R. leguminosarum bv. viciae 3841 induces the formation of nitrogen-fixing nodules with several hosts, including pea (Pisum sativum), lentil (Lens culinaris), and Vicia spp. [74]. As it is common in rhizobia, the Rlv3841 nodD gene, which is transcribed divergently from the nodABCIJ operon, is constitutively expressed, although its expression is negatively autoregulated [75,76]. As commented previously, the exhaustive transcriptomic study performed by Ramachandran et al. [42] was focused on the understanding of the adaptation of R. leguminosarum bv. viciae 3841 to pea, alfalfa and sugar beet rhizospheres in order to establish the sets of genes involved in the general adaptation to the rhizosphere and in the specific adaptation to the pea rhizosphere. In this study, it was also analysed the transcriptomic response to the presence of hesperetin, a nod gene-inducing flavonoid of this bacterium. The analysis of these data indicate that 33 genes are differentially expressed (27 upregulated and 6 repressed) upon flavonoid treatment (Table S2), which are distributed between two groups: (i) genes controlled by NB and (ii) genes not preceded by a NB (Figure 3).

Figure 3

Transcriptomic response in the presence of hesperetin in R. leguminosarum bv. viciae 3841. Proteins are illustrated by blue circles. Activation and repression of genes are indicated by black and red arrows, respectively. Genes regulated by a nod box are represented in green. Genes which are not controlled by a nod box are indicated in grey.

3.2.1. Genes Controlled by NB

The presence of at least 5 NB controlling their respective set of nod genes had been previously identified in the genome R. leguminosarum bv. viciae 3841 [77,78,79,80,81,82,83]. As expected, the expression data upon flavonoid induction indicate that the transcriptional units controlled by NB (nodABCIJ, nodFEL, nodMNT, and nodO) in this bacterium are strongly upregulated. This set of nod genes encoded proteins implied in the synthesis and export of the pool of NF produced by Rlv3841 [42] (Table S2) (Figure 3).

3.2.2. Genes Not Preceded by a NB

One of the most abundant proteins produced by strains of Rhizobium leguminosarum bv. viciae in the rhizosphere is RhiA [84]. The rhiA gene is part of the rhiABC operon that is transcriptionally controlled by the LuxR-type regulator encoded by rhiR. It has been previously described that these genes are strongly activated in the rhizosphere of plants via inducing flavonoids. The functions of the proteins encoded by this operon seems to be related with the plant–microbe interaction, possibly by allowing the bacteria to metabolize some plant-made metabolites [85,86]. As expected, the transcriptomic study indicates that the rhiABC-rhiR gene cluster is highly activated in the presence of hesperetin. This set of genes is located between nitrogen fixation (nifHDK) and nodulation (nod) genes on the symbiotic plasmid, pRL10, but are not dependent on a NB. Interestingly, the matBC genes are also upregulated in the presence of hesperetin. The matABC genes code for a malonyl-CoA decarboxylase, malonyl-CoA synthetase, and a putative malonate transporter, respectively [87,88]. These enzymes are involved in the uptake and conversion of malonate to acetyl-CoA. Again, the upregulation of these genes seems to be orientated towards rhizobial adaptation when the bacterium colonizes the legume root, since the level of free malonate in legumes is known to be increased under symbiotic conditions [89]. Interestingly, in pea roots, the uptake and catabolism of malonate plays an important role in the initial rhizobial attachment and colonization of the root [90]. Thus, transcriptomic data suggest that flavonoid perception and malonate uptake is co-regulated in R. leguminosarum bv. viciae 3841, probably in order to ensure the attachment to the root surface of a compatible host plant. Lastly, 6 genes, including flaD, lpxC and aldA (encodes for flagellin, N-acetylglucosamine deacetylase and alanine dehydrogenase, respectively), and others coding for proteins with unknown functions (RL3912, RL0798, and RL3172), are downregulated in Rlv3841 in the presence of inducing flavonoids but their relevance in symbiosis requires further research (Table S2) (Figure 3).

3.3. Rhizobium Tropici CIAT 899: NF Synthesis Also under Stressing Conditions, Production of a Large Variety of NF and Phytohormones

Rhizobium tropici CIAT 899 is a broad host-range bacterium that effectively nodulates several legumes, including Phaseolus vulgaris, Leucaena leucocephala, Lotus japonicus and L. burtii [91,92,93]. The main characteristics of this strain include its high tolerance to several environmental stresses such as high temperature, acidity or salinity and its capacity to produce a large variety of NF in the presence of inducing flavonoids, such as apigenin [91,94,95]. Surprisingly, under salt stressing conditions, this bacterium induces the synthesis of NF in a NodD1-independent manner [96,97]. Interestingly, genome sequencing of R. tropici CIAT 899 revealed the presence of five different nodD genes and three different nodA genes in the symbiotic plasmid [98]. The nodA1 gene is located close to nodD1, whose encoded protein is the key regulator of NF synthesis upon induction with flavonoids [92,99] and together with nodBC comprise the operon responsible for the synthesis of the NF core. The nodA2 gene, which is adjacent to nodD2, is part of a set of genes, including hsnT and nodFE, implied in unsaturated fatty acid incorporation into NF molecules. Finally, the nodA3 gene is located downstream the nodD3 gene but no other symbiotic-related genes have been identified in its vicinity [93].

Recently, Pérez-Montaño et al. [44] performed the first transcriptomic study in R. tropici CIAT 899 in the presence of apigenin and salt. This work revealed that 19 genes are differentially expressed when the bacterium was grown in the presence of apigenin, being 15 genes upregulated and located on the symbiotic plasmid, meanwhile the other 4 are downregulated and distributed in other replicons. Interestingly, with respect to the cultures supplemented with salt, the same symbiotic-related genes are also upregulated, but many other non-symbiotic genes are also differentially expressed (95 upregulated and 628 downregulated) (Table S3). Interestingly, in a second transcriptomic approach, del Cerro et al. [45] confirmed that the transcriptional activation of previously mentioned genes upon apigenin and salt conditions are mediated by NodD1 and NodD2, respectively. Taking in consideration both transcriptomic studies, two groups of genes are identified: (i) genes controlled by nod boxes (NB); and (ii) genes not preceded by a NB (Figure 4).

Figure 4

Transcriptomic response in the presence of apigenin and NaCl in R. tropici 899. Proteins are illustrated by blue circles. The DNA to protein translation process is represented by dotted arrows. Activation and repression of genes are indicated by black and red arrows, respectively. Genes regulated by a nod box are represented in green.

3.3.1. Genes Controlled by NB

Four sets of genes are activated in the symbiotic plasmid (NC_020061) in a NodD1-apigenin dependent manner and/or in a NodD2-salt dependent manner. Two of them corresponded to the nodA1BCSUIJHPQ1Q2 and nodA2hsnTnodFE operons (NB1 and NB2, respectively), the third is involved in the synthesis of the phytohormone indole-3-acetic acid (IAA) (NB4), and the last set of genes (PB01550 and PB01545) codes for proteins with unknown functions (NB5) [44]. Although nodM expression did not reach the threshold considered by the authors of the study (fold change ±4), qRT-PCR (quantitative Real Time-Polymerase Chain Reaction) results indicated that this gene is also significantly expressed in the presence of the flavonoid apigenin [44]. Therefore, at least two main biological processes that are beneficial for symbiosis are activated in this bacterium when induced with apigenin or salt: the synthesis and export of NF, that will induce root nodule primordia formation (nodA1BCSUIJHPQ1Q2, nodA2hsnTnodFE, and nodM) and the production of indole-3-acetic acid (IAA) that will favour root development (y4wEF). Interestingly, the PB01055 gene, which is located on the symbiotic plasmid under the control of NB7 and codes for a putative calcium/calmodulin dependent protein kinase II, is only transcriptionally activated in the presence of salt. As commented above, the transcriptomic analyses indicate that NodD1 is not involved in the transcriptional activation of symbiotic genes under salt stress. However, there is an exception, since transcription of the nodD2 gene is controlled by NB9 in a salt-NodD1 dependent manner. This finding indicates that NodD1 could increase nodD2 expression and, consequently, the production of NF in the presence of high concentrations of salt (Table S3) (Figure 4).

3.3.2. Genes Not Preceded by a NB

According to the transcriptomic study carried out by Pérez-Montaño et al. [44], 4 genes are differentially repressed in a NB-independent manner in the presence of inducing flavonoids. These genes are located in different replicons and code for conjugal Transfer protein (TraG) (located on the 0.22-Mb pRtrCIAT899a), 2 ABC-type transporters (one on the 3.8-Mb chromosome and the other on the 2.08-Mb pRtrCIAT899c) and an oxidoreductase (located on pRtrCIAT899c). However, so far, no reports have been published describing the implication of any of these genes in the rhizobium–legume symbiosis (Table S3) (Figure 4). Interestingly, it has been recently described that TraG is critical for promoting the transference of conjugant elements from Azorhizobium caulinodans to M. huakuii in the rhizosphere [100]. Thus, the up-regulation of the R. tropici CIAT 899 traG gene in response to apigenin might indicate that rhizobia, when sense the presence of a putative host legume plant, could transfer by conjugation its symbiotic plasmid to related bacteria.

3.4. Sinorhizobium Meliloti 1021: Only Production and Export of NF?

S. meliloti 1021 is a soil bacterium able to induce the formation of nodules in a very reduced number of legume genera: Medicago, Melilotus and Trigonella, including Medicago sativa (alfalfa) and the model legume Medicago truncatula. In S. meliloti, the expression of the nodulation genes, nod, nol, and noe operons, can be induced by three different NodD proteins: NodD1 and NodD2, which are activated by specific flavonoids and expressed constitutively, and NodD3, whose expression is subject to a complex regulation involving the regulatory proteins SyrM and NodD1, and the nitrogen status of the cell [101,102]. Although inactivation of the three nodD genes is required to completely abolish nodulation, the symbiotic relevance of each of these regulators depends on the host legume. Thus, the main regulator for nodulation with Medicago sativa is NodD1 [101].

The tripartite genome of S. meliloti 1021 comprises a 3.65-Mb chromosome and the 1.35-Mb pSymA and 1.68-Mb pSymB megaplasmids [103]. Many transcriptomics reports relevant to symbiosis have been previously published for this bacterium (see Table 1). In two of them, the transcriptomic responses in the presence of the inducing flavonoid luteolin were assessed [38,46]. In these reports, 24 genes differentially expressed were identified upon flavonoid treatment (Table S4), 18 of them belonging to the nod/noe/nol operons and 6 lacking a NB. However, high variations among replicates were scored by the authors in the expression of several of the genes belonging to the latter group, suggesting that maybe they are not actually differentially expressed genes upon treatment with luteolin. In any case, these transcriptomic studies unequivocally identified two different groups of genes depending on their promoter sequences: (i) genes controlled by NB and (ii) genes not preceded by a NB (Table S4).

3.4.1. Genes Controlled by NB via NodD1

Eighteen upregulated genes controlled by six NB were identified on the pSymA of S. meliloti 1021. Five canonical nod boxes are located upstream of the nodABC, nodFEGP1P2, nodH, nodLnoeAB and nodMnolFG operons and another less conserved nod-box is situated upstream of syrM [46]. The proteins encoded by genes upregulated in the presence of luteolin via NB are implied in the synthesis and export of NF, with the exception of syrM and two not previously described genes (SMa0850 and SMa0848).

3.4.2. Genes Not Preceded by a NB

The absence of nod boxes in the promoter regions of the differentially expressed genes SMc03151, SMc03167, SMc03168, SMc03169, and groEL5 suggests a NodD1-indirect mechanism of activation in the presence of flavonoids. The SyrM protein might be responsible for the activation of this set of genes. However, no syrM boxes have been found in the promotor regions of these differentially expressed genes [46]. Independently of their regulation, the functions of the proteins encoded by these genes are currently unknown or poorly investigated, and for this reason, further research is needed to advance in the knowledge about the role of these proteins in the symbiosis between S. meliloti 1021 and its legume hosts.

3.5. Sinorhizobium Fredii HH103: A Hierarchical Regulatory Cascade Controls NF Production and Export, Biosynthesis of Phytohormones, and Assembly of the T3SS

Sinorhizobium fredii is a rhizobial species characterized by its extremely broad host range since it is capable of nodulating more than 100 genera of legumes including plants forming determinate (such as Glycine max) and indeterminate nodules (such as Glycyrrhiza uralensis) [104,105,106]. Different S. fredii strains has been described so far, being the four most studied NGR234, USDA257, CCBAU45436, and HH103. In fact, genomic information is available for all of them [107,108,109,110,111]. An interesting difference between these four strains resides in their symbiotic behaviour with soybean (G. max); NGR234 cannot nodulate this legume [104], USDA257 and CCBAU45436 induce the formation of nodules only on wild and non-commercial soybeans [112], but HH103 nodulates effectively both non-commercial and commercial American soybeans [113].

S. fredii HH103 strain is one of the most studied rhizobial strains, since its genome has been sequenced and the symbiotic capacity of a large number of mutants affected in Nod-factor production, bacterial surface polysaccharides, and the T3SS have been determined [106,109]. The S. fredii HH103 genome is composed of a chromosome and six plasmids. The symbiotic plasmid of HH103 harbours, besides to the structural nod genes, diverse symbiotic regulatory genes, including syrM, ttsI, and two copies of nodD. In order to know the global expression response of S. fredii HH103 to the presence of flavonoids, Pérez-Montaño et al. [47] performed transcriptomic studies in the presence and absence of genistein, an inducing flavonoid of the nod genes in this strain. In addition, nodD1 and ttsI mutants were also included in this study to determine the role of NodD1 and TtsI in the regulation of the HH103 genes differentially expressed by genistein.

In this transcriptomic report, 100 differentially expressed genes were identified upon genistein treatment (Table S5). Thus, 92 genes were upregulated, most of them located on the symbiotic plasmid (psfHH103d), and only 8 genes were repressed under these conditions. Interestingly, transcriptomic studies carried out with the nodD1 and ttsI mutants unequivocally identified three different groups of genes depending on the presence of functional promoter sequences and on the transcriptional regulator: (i) genes controlled by NB via NodD1; (ii) genes regulated by TB via TtsI; and (iii) genes not preceded by a NB or a TB (Figure 5).

Figure 5

Transcriptomic response in the presence of genistein in S. fredii HH103. Proteins are illustrated by blue circles. The DNA to protein translation process is represented by dotted arrows. Activation and repression of genes are indicated by black and red arrows, respectively. Genes regulated by a nod box or a tts box are represented in green and purple, respectively. Genes which are not controlled by a nod box or a tts box are indicated in grey. NB: nod box; TB: tts box; SB?: genes that possibly are controlled by a syrM box.

3.5.1. Genes Controlled by NB via NodD1

Fifteen potential NB have been located on the symbiotic plasmid of S. fredii HH103 [109]. The transcriptomic comparison between the activation patterns of the wild-type and nodD1 mutant strains defined 11 of the 15 previously described NB as active (NB1, NB2, NB8, NB9, NB10, NB13, NB14, NB15, NB17, NB18 and NB19). The functions of the proteins encoded by these genes include synthesis and transport of NF (nodZnoeLnolK and nodABCIJnolO’noeI operons), electron transfer to nitrogenase (fixABCX genes), synthesis of the phytohormone IAA (y4wE gene), putative production of hopanoids (hpn cluster) and transcriptional regulators such as TtsI and SyrM. Interestingly, 4 sets of genes (psfHH103d_116-118, psfHH103d_161, psfHH103d_447-448 and psfHH103d_208), also controlled via NB, NodD1 and flavonoids, code for conserved hypothetical proteins whose functions are currently unknown or poorly investigated [47] (Table S4) (Figure 5). Interestingly, one of these proteins, encoded by psfHH103d_161, could be related to the negative regulation of exopolysaccharide (EPS) production by NodD1 and inducing flavonoids described for S. fredii HH103 [31]. This hypothetical protein is 58% identical to R. leguminosarum biovar phaseoli PsiB, a protein involved in exopolysaccharide inhibition [114]. Further research is needed to clarify this issue.

3.5.2. Genes Controlled by TB via TtsI

Some Gram-negative bacteria possess a specialized apparatus for protein secretion called T3SS [115]. Some pathogenic and symbiotic bacteria, including rhizobia, use this secretion system to deliver effector proteins into the eukaryotic cell to interfere with signal transduction pathways and promote infection by suppressing defence responses [116]. The transcriptomic study performed in S. fredii HH103 corroborated that its symbiotic T3SS is produced and assembled in a flavonoid-, NodD1-, and TtsI-dependent manner, since the ttsI gene is located downstream of a functional NB [47]. In addition, this study determined than 11 out of the 18 previously defined as potential TB were functional and controlled the expression of 35 ORFs. All these TB were located on the symbiotic plasmid (psfHH103d) with the exception of TB1 and TB2, present in plasmid psfHH103c. Most of the genes upregulated via TB and TtsI code for components of the T3SS apparatus (nopBrhcJnolUrhcLNQRSTU and nopCAy4yQ) and other proteins putatively involved in the translocation of proteins to the cytoplasm of the plant cell, such as NopX and GunA. The rest of genes controlled by the remaining TB code for effector proteins, namely nopC, nopD, nopI, nopM, nopM2, nopL, nopP, and nopT (Table S4) (Figure 5). In fact, many of the proteins encoded by these genes were previously detected in supernatants of S. fredii HH103 obtained from cultures induced with genistein [117,118]. These findings indicate that, in the presence of inducing flavonoids, this bacterium produce a fully functional T3SS.

3.5.3. Genes Not Preceded by a NB or a TB

This last group of genes comprises genes differentially expressed in the presence of genistein but not located downstream of a NB or a TB. The transcriptomic analysis indicates that 30 genes distributed throughout the chromosome and plasmids are differentially expressed, 22 of them induced and only 8 repressed when S. fredii HH103 is cultured in the presence of genistein (Table S4 and Figure 5) [47]. Among the differentially expressed genes, two operons (psfHH103d_322-psfHH103d_319 and psfHH103d_306-psfHH103_d311), which code for proteins not previously described, seems to be controlled by the regulator SyrM (activated by flavonoids in a NodD1- and NB-dependent manner), since potential syrM boxes (SB) has been identified upstream of these genes [47]. On the other hand, seven genes (SFHH103_00346-00348, SFHH103_00844, SFHH103_01920 and psfHH103d_274-275) that also encode for hypothetical proteins, seem to be controlled by both NodD1 and TtsI, but no promoter regions have been identified in the sequences preceding the ORFs. Lastly, the expression of four genes (the SFHH103_03875 gene and the SFHH103_05321-SFHH103_05319 operon) is affected in the presence of genistein in a NodD1- and TtsI-independent manner (Table S4) (Figure 5). The chromosomic SFHH103_03875 gene, which is induced with genistein, codes for a putative membrane permease containing two EamA domains. Despite of this gene is well conserved among different rhizobia, it has not been characterized so far. In contrast, the SFHH103_05321-SFHH103_05319 operon, which codes for a TetR transcriptional regulator and two multidrug efflux transporter proteins, is repressed in the presence of inducing flavonoids. Interestingly, several efflux transporter systems are also differentially expressed in other soybean symbiont, B. japonicum USDA 110, in the presence of genistein [43], but further investigations are needed to shed light about the putative role of these transport systems in symbiosis.

4. Conclusions

The understanding of the molecular mechanisms governing the rhizobia–legume symbiosis is crucial for the improvement of this interaction and, consequently, for increasing the yield of crops with high agronomical importance. This process involves an interchange of symbiotic signals between both partners. It is well known that the rhizobial protein NodD is activated by specific flavonoids exuded by legume roots and then binds to conserved promoter sequences activating the transcription of the nod genes, which encode for enzymes responsible for the biosynthesis and secretion of the NF. Interestingly, many other symbiotic-related traits are regulated in first instance via flavonoids. In this review, an exhaustive compilation of many symbiosis-related transcriptomic analysis (microarray and RNA-seq) carried out in rhizobia during the last two decades is summarized. Our attention has been focused on the reports describing the transcriptomic changes of several rhizobia triggered upon flavonoid treatment in order to know what are the arsenals of weapons that escorts NF in different rhizobia to successfully infect host legumes. The global transcriptomic reports analysed in this paper indicate that each bacterium has developed specific molecular responses to achieve this purpose (Figure 6). However, the possibility that other processes could be triggered in the presence of inducing flavonoids cannot be ruled out, since the RNA-seq and microarrays techniques commonly give a single frame of the movie. Lastly, further efforts must be performed in order to characterize those differentially expressed genes whose function remains unknown, as well as for the analysis of transcriptomic changes mediated by flavonoids in not previously studied rhizobia. These studies will improve our understanding of the molecular strategies developed by each bacterium to success in the nodulation process and, therefore, shed light on the molecular dialogue established during the symbiotic interaction.

Figure 6

Molecular responses triggered by inducing flavonoids according to transcriptomic reports in several rhizobial strains. NB: nod boxes, TB: tts boxes. Black arrows: activation. Red Arrows: repression. Dotted arrows indicate transport across the membrane and cell wall.